Statistical evaluation

The data collected were analysed for significant patterns and trends. The statistical analyses w ere made using the SAS software 6.12 (Anon. 1990). General Linear M odels (GLM) procedure was used to evaluate data For frequency data, the chi-square test w as used to evaluate differences betw een groups.

2.3 Results and discussion

2.3.1 Quality o f the wall studs

The studs were graded according to the requirements shown in Table 5. Those exceeding th ese lim its for any one type o f warp were classed as out-of-grade.

Table 5. Summary of proposed requirements on dimension, warp and moisture content on wall studs according to “Guidelines for Purchasing Building Timber” (Johansson et al. 1993).

DIMENSION

Width (mm) Thickness (mm)

45 ±1 70, 95, 120, 145, 170 ± 1

SHAPE - limits are set for the whole stud or for 3 m length

Cup Bow Crook Twist

max 2% of width max 6 mm max 4 mm max 4% of width,

5 mm is allowed MOISTURE CONTENT

1) The timber should be delivered kiln-dried to a moisture content of 15 ± 2%, alternatively at moisture content class 12 (i.e. 9-15%) according to SS232740 (Anon. 1991).

2) Moisture content of 18 + 2%, maximum 22%, could be accepted i f the warp requirements when dried to equilibrium moisture content are fulfilled and guaranteed by the supplier.

D im ensions

T he requirements for dim ensions fall into the category o f functional requirements.

According to Johansson et al. (1994b), some contractors claim that the dim ensions o f the products bought are smaller than those exp ected by the standards. H ow ever, the end-users usually have no serious problems with the dim ensions.

T he necessary thickness is determined from the strength and stiffness requirements and from the need for sufficient support for sp licin g the cladding boards. The width should not differ between or within studs, i.e., the wall should not become bent (Johansson et al. 1994b).

According to the requirements shown in Table 5, width and thickness are not allowed to deviate more than ± 1 mm from the specified measurements. On average, 94% o f the studs in the sample packages met this requirement. However, in one o f the packages the requirements were fulfilled by only 66% o f the studs. The result for the other packages varied betw een 91 and 100% accepted studs (Figure 5). The reason for rejection o f studs was that the measures were lower than the minimum value for width or thickness.

Figure 5. Percentage of the studs in the sample packages from each sawmill that was accepted with regard to dimensions (width x thickness).

M oisture content

The moisture content itself is not an important measure for the user, but many o f the w ood properties change as moisture content decreases or increases. If moisture content is too high at the time of building, studs will warp when drying to the equilibrium moisture content (EMC) and the wall risks becoming bent. The best way to avoid these problems is to deliver timber with the correct EMC. H ow ever, a prerequisite is that the w ood is handled and stored in such a way that it is not exposed to moisture at the building site.

T oday, sawn timber is usually dried to a moisture content class 18, i.e. 18 ± 6% (Anon.

1991). This limit is set primarily to avoid attack by microorganisms during transportation.

T he EM C for interior wall studs in Sweden varies over the year from approximately 8% in the winter to 13% in the summer (Bergstrom 1981). The requirement for moisture content o f the sawn timber at time o f delivery proposed in Guidelines for Purchasing B uilding timber (Johansson et al. 1993) is 15 ± 2%. A maximum moisture content o f 20% is accepted

if

Figure 6 shows the distribution o f moisture content o f the boards in the sample packages.

T he frequency distribution shows that the target moisture content, 18%, is rather w ell reached by the sawmills. Average moisture content was 18% for three o f the packages, 16% for two packages and 19% for one package.

40

12 13 14 15 16 17 18 19 20 21 22

Moisture content (%)

Figure 6. Frequency distribution of moisture content after planing for studs in the sample packages.

A lm ost all (91%) studs had a moisture content o f 20% or less after planing. However, in one package only 46% o f the studs were accepted. T he variation in moisture content am ong the rejected studs in that package was low and ranged between 20.1-21.5% . In the other packages between 93 and 100% o f the studs were accepted.

If the limit 15±2% moisture content was used instead, the variation between packages w as statistically significant and m uch larger. In four packages, less than 20% o f the studs fulfilled the requirement of a moisture content between 13 and 17%. The remaining tw o packages had 74 and 87% studs that were accepted.

Table 6. Percentage of studs in each sample package that was accepted with regard to the requirements on moisture content.

Sawmill Package Percentage (%) of studs accepted with regard to moisture content

<20 15 ± 2

A 1 46 0

A 2 95 1

B 3 97 87

C 4 100 12

D 5 93 17

E 6 100 74

Warp

T h e proposed grading rules, show n in table 5, are valid on dried and planed studs.

To be fully accepted the studs are not allowed to exceed any o f the warp lim its. S in ce the saw m ills usually dry structural timber to a higher moisture content than that recommended in the Guidelines for Purchasing Building Timber (Johansson et al. 1993), moisture content was not considered in the follow in g evaluation of warp.

O f the total number o f studs in this study, i.e. sam ple studs after planing and delivery studs, 66% fulfilled the requirements for all four types of warp. This means that as much as approximately one-third o f the studs graded,

did not

According to the interview study by Johansson et al. (1994a and b) tw ist is the m ost prevalent problem in building timber today. Also, crook is considered as a serious problem, w hereas bow often can be adjusted during erection o f the wall. Figure 7 sh ow s the percentage o f studs accepted or rejected for each type o f warp.

Twist Crook Bow Warp

Figure 7. Percentage of studs accepted or rejected for each type of warp and for the combination of twist, crook and bow (warp).

W hen only twist was considered, about 80% o f all studs in the material fu lfilled the requirement o f maximum 5 mm twist. According to Figure 7, 92% o f the total number o f studs fulfilled the requirement of maximum 4 mm crook. B ow was less than 6 mm in 87% o f the studs.

M axim um cup allowed is set to 2% o f the width o f the stud, i.e. 1.9, 2 .4 and 2.9 mm for the 95-, 120- and 145-mm w ide studs, respectively. In this study, the largest cup measured was 1 mm and this only for few of the studs. Therefore, cup was not considered a problem for the dimensions studied and will not be discussed further in this chapter.

Comparison o f individual studs

Figure 8 shows the types o f warp o f individual studs. Only a few studs ex ceed ed the

Twist Bow Crook Twist Twist Crook Twist,

and and and Crook

Bow Crook Bow and

Bow Figure 8. Number of warp defects exceeding the maximum limit, per stud in both the sample and the delivery packages.

The m ost com m on single form o f warp was twist, followed by bow and crook. When two types o f warp were present in one single stud the combination o f tw ist and bow was m ost frequent. Crook and bow appeared together in only 1% o f the studs.

Comparison o f saw m ills and packages

A comparison o f the sawmills showed that there were large differences between sawmills concerning the percentage o f studs that fulfilled the proposed limits for warp (Table

7).

Table 7. Percentage of studs in each sample and delivery package that was accepted with regard to the requirements on twist, crook and bow and the percentage of studs in each package that fulfilled all the requirements on warp. S=sample package, D=delivery package.

Sawmill Package number

Percentage (%) of studs accepted with regard to

Twist Crook Bow Warp

A 1 (S) 96 91 92 80

2(S ) 80 94 78 61

B 3 (S) 72 88 77 48

7 (D ) 58 89 83 42

C 4 (S ) 87 86 95 71

8(D ) 90 96 93 81

D 5 (S) 72 62 66 34

9 (D ) 90 93 89 76

E 6 (S) 65 96 77 47

10(D) 87 95 95 79

A t the best sawm ill (C), 78% o f the total number o f studs fulfilled the requirements for warp. Only 44% o f the studs from sawmill B were accepted, whereas sawm ills A, D and E had 65-71% accepted studs in their packages.

In the best package, 81% o f the studs were accepted whereas in the package with the low est quality, only 34% o f the studs were accepted when twist, crook and bow were considered together.

A pairwise comparison of the packages from each sawmill (Figure 9) shows that there were also differences between packages within the same sawmill.

-oV.

to3 T3D

&, uo O

a

D

£ 100

80 60 40 20 0

C A E D B

■ Sample

□ Delivery

Sawmill

Figure 9. Comparison of the percentage of accepted studs in the sample and delivery packages from each sawmill.

D elivery packages from saw m ills C, D and E had significantly more accepted studs than the sample packages. T he opposite was true for sawmill B. A lso, at sawm ill A, where both packages were sampled at the same step in the production chain, there was a significant difference in percentage of accepted studs between the packages.

Both Table 7 and Figure 9 show that there was a significant difference in yield between sawm ills and between packages within the sawmills. Large variation in warp between packages has also been noted in an evaluation o f stud quality at two building sites in Sw eden (Lindvall 1996). T his variation could probably be attributed to differences in raw material, as well as processing factors.